Instance variable pull up broken
Locked
Instance variable pull up broken
 
|
Here's a funny way to crash your image. Works in:
Pharo-1.1-11367-Beta Latest update: #11367 Make a class named AbstractBla, with two subclasses: SubA, SubB. Then, add to both SubA and SubB the identically named instance variable: instVar. Now, use the "pull up" refactoring in OmniBrowser on SubA to pull up instVar into the superclass. As you correctly guess, that breaks SubB, because now both its superclass and SubB have an instance variable named instVar. But, instead of refusing to refactor, OmniBrowser dutifully refactors and takes the image with it. I hope that was understandable. Cheers, Niko -- http://scg.unibe.ch/staff/Schwarz twitter.com/nes1983 Tel: +41 076 235 8683 _______________________________________________ Pharo-project mailing list [hidden email] http://lists.gforge.inria.fr/cgi-bin/mailman/listinfo/pharo-project |
Re: Instance variable pull up broken
|
|
I cannot reproduce. If you look at the code you should see that what
you describe should not happen: transform class allSubclasses do: [:each | (each directlyDefinesInstanceVariable: variableName) ifTrue: [each removeInstanceVariable: variableName]]. class addInstanceVariable: variableName Maybe your class hierarchy is broken? Lukas On 7 August 2010 22:18, Niko Schwarz <[hidden email]> wrote: > Here's a funny way to crash your image. Works in: > > Pharo-1.1-11367-Beta > Latest update: #11367 > > Make a class named AbstractBla, with two subclasses: SubA, SubB. Then, > add to both SubA and SubB the identically named instance variable: > instVar. Now, use the "pull up" refactoring in OmniBrowser on SubA to > pull up instVar into the superclass. As you correctly guess, that > breaks SubB, because now both its superclass and SubB have an instance > variable named instVar. But, instead of refusing to refactor, > OmniBrowser dutifully refactors and takes the image with it. > > I hope that was understandable. > > Cheers, > > Niko > > > -- > http://scg.unibe.ch/staff/Schwarz > twitter.com/nes1983 > Tel: +41 076 235 8683 > > _______________________________________________ > Pharo-project mailing list > [hidden email] > http://lists.gforge.inria.fr/cgi-bin/mailman/listinfo/pharo-project > -- Lukas Renggli www.lukas-renggli.ch _______________________________________________ Pharo-project mailing list [hidden email] http://lists.gforge.inria.fr/cgi-bin/mailman/listinfo/pharo-project |
Re: Instance variable pull up broken
|
|
Hmm, I can't reproduce it in a fresh image, either. Sorry for the noise.
Niko On Sat, Aug 7, 2010 at 10:39 PM, Lukas Renggli <[hidden email]> wrote: > I cannot reproduce. If you look at the code you should see that what > you describe should not happen: > > transform > class allSubclasses do: > [:each | > (each directlyDefinesInstanceVariable: variableName) > ifTrue: [each removeInstanceVariable: variableName]]. > class addInstanceVariable: variableName > > Maybe your class hierarchy is broken? > > Lukas > > On 7 August 2010 22:18, Niko Schwarz <[hidden email]> wrote: >> Here's a funny way to crash your image. Works in: >> >> Pharo-1.1-11367-Beta >> Latest update: #11367 >> >> Make a class named AbstractBla, with two subclasses: SubA, SubB. Then, >> add to both SubA and SubB the identically named instance variable: >> instVar. Now, use the "pull up" refactoring in OmniBrowser on SubA to >> pull up instVar into the superclass. As you correctly guess, that >> breaks SubB, because now both its superclass and SubB have an instance >> variable named instVar. But, instead of refusing to refactor, >> OmniBrowser dutifully refactors and takes the image with it. >> >> I hope that was understandable. >> >> Cheers, >> >> Niko >> >> >> -- >> http://scg.unibe.ch/staff/Schwarz >> twitter.com/nes1983 >> Tel: +41 076 235 8683 >> >> _______________________________________________ >> Pharo-project mailing list >> [hidden email] >> http://lists.gforge.inria.fr/cgi-bin/mailman/listinfo/pharo-project >> > > > > -- > Lukas Renggli > www.lukas-renggli.ch > > _______________________________________________ > Pharo-project mailing list > [hidden email] > http://lists.gforge.inria.fr/cgi-bin/mailman/listinfo/pharo-project > -- http://scg.unibe.ch/staff/Schwarz twitter.com/nes1983 Tel: +41 076 235 8683 _______________________________________________ Pharo-project mailing list [hidden email] http://lists.gforge.inria.fr/cgi-bin/mailman/listinfo/pharo-project |
Re: Instance variable pull up broken
|
|
if you can reproduce it or understand what was your problem we are interested because we
want to make sure that the next classbuilder should be really robust. Stef On Aug 8, 2010, at 1:24 AM, Niko Schwarz wrote: > Hmm, I can't reproduce it in a fresh image, either. Sorry for the noise. > > Niko > > On Sat, Aug 7, 2010 at 10:39 PM, Lukas Renggli <[hidden email]> wrote: >> I cannot reproduce. If you look at the code you should see that what >> you describe should not happen: >> >> transform >> class allSubclasses do: >> [:each | >> (each directlyDefinesInstanceVariable: variableName) >> ifTrue: [each removeInstanceVariable: variableName]]. >> class addInstanceVariable: variableName >> >> Maybe your class hierarchy is broken? >> >> Lukas >> >> On 7 August 2010 22:18, Niko Schwarz <[hidden email]> wrote: >>> Here's a funny way to crash your image. Works in: >>> >>> Pharo-1.1-11367-Beta >>> Latest update: #11367 >>> >>> Make a class named AbstractBla, with two subclasses: SubA, SubB. Then, >>> add to both SubA and SubB the identically named instance variable: >>> instVar. Now, use the "pull up" refactoring in OmniBrowser on SubA to >>> pull up instVar into the superclass. As you correctly guess, that >>> breaks SubB, because now both its superclass and SubB have an instance >>> variable named instVar. But, instead of refusing to refactor, >>> OmniBrowser dutifully refactors and takes the image with it. >>> >>> I hope that was understandable. >>> >>> Cheers, >>> >>> Niko >>> >>> >>> -- >>> http://scg.unibe.ch/staff/Schwarz >>> twitter.com/nes1983 >>> Tel: +41 076 235 8683 >>> >>> _______________________________________________ >>> Pharo-project mailing list >>> [hidden email] >>> http://lists.gforge.inria.fr/cgi-bin/mailman/listinfo/pharo-project >>> >> >> >> >> -- >> Lukas Renggli >> www.lukas-renggli.ch >> >> _______________________________________________ >> Pharo-project mailing list >> [hidden email] >> http://lists.gforge.inria.fr/cgi-bin/mailman/listinfo/pharo-project >> > > > > -- > http://scg.unibe.ch/staff/Schwarz > twitter.com/nes1983 > Tel: +41 076 235 8683 > > _______________________________________________ > Pharo-project mailing list > [hidden email] > http://lists.gforge.inria.fr/cgi-bin/mailman/listinfo/pharo-project _______________________________________________ Pharo-project mailing list [hidden email] http://lists.gforge.inria.fr/cgi-bin/mailman/listinfo/pharo-project |
Re: Instance variable pull up broken
|
|
In reply to this post by Lukas Renggli
I've been trying out the refactoring tools with a specific 2
subclasses of a common class example - and I didn't get this problem either - I could push up methods and pull up instance variables of my two common subclasses. The only point I would note - it seems inconsistent that you "pull up" instance variables but "push up" methods (I recall it was a naming act of history, but it would be nice to call it "push up" everywhere.) Tim On 7 Aug 2010, at 21:39, Lukas Renggli wrote: > I cannot reproduce. If you look at the code you should see that what > you describe should not happen: > > transform > class allSubclasses do: > [:each | > (each directlyDefinesInstanceVariable: variableName) > ifTrue: [each removeInstanceVariable: variableName]]. > class addInstanceVariable: variableName > > Maybe your class hierarchy is broken? > > Lukas > > On 7 August 2010 22:18, Niko Schwarz <[hidden email]> > wrote: >> Here's a funny way to crash your image. Works in: >> >> Pharo-1.1-11367-Beta >> Latest update: #11367 >> >> Make a class named AbstractBla, with two subclasses: SubA, SubB. >> Then, >> add to both SubA and SubB the identically named instance variable: >> instVar. Now, use the "pull up" refactoring in OmniBrowser on SubA to >> pull up instVar into the superclass. As you correctly guess, that >> breaks SubB, because now both its superclass and SubB have an >> instance >> variable named instVar. But, instead of refusing to refactor, >> OmniBrowser dutifully refactors and takes the image with it. >> >> I hope that was understandable. >> >> Cheers, >> >> Niko >> >> >> -- >> http://scg.unibe.ch/staff/Schwarz >> twitter.com/nes1983 >> Tel: +41 076 235 8683 >> >> _______________________________________________ >> Pharo-project mailing list >> [hidden email] >> http://lists.gforge.inria.fr/cgi-bin/mailman/listinfo/pharo-project >> > > > > -- > Lukas Renggli > www.lukas-renggli.ch > > _______________________________________________ > Pharo-project mailing list > [hidden email] > http://lists.gforge.inria.fr/cgi-bin/mailman/listinfo/pharo-project _______________________________________________ Pharo-project mailing list [hidden email] http://lists.gforge.inria.fr/cgi-bin/mailman/listinfo/pharo-project |
Re: Instance variable pull up broken
|
|
> The only point I would note - it seems inconsistent that you "pull up"
> instance variables but "push up" methods (I recall it was a naming act of > history, but it would be nice to call it "push up" everywhere.) That was unified a few months ago. Lukas -- Lukas Renggli www.lukas-renggli.ch _______________________________________________ Pharo-project mailing list [hidden email] http://lists.gforge.inria.fr/cgi-bin/mailman/listinfo/pharo-project |
«
Return to Pharo Smalltalk Developers
|
1 view|%1 views
| Free forum by Nabble | Edit this page |